THE EFFECT OF TEST LENGTH AND IRT MODEL ON THE DISTRIBUTION AND STABILITY OF 3 APPROPRIATENESS INDEXES

The extent to which three appropriateness indexes-Z3, ECIZ4, and W (a variation of Wright's person-fit statistic)-are well-standardized was investigated in a monte carlo study. To assess the effects of the item response theory (IRT) model and test length on the distribution of th e indexes and their cutoff values at three false positive rates, nonab errant response patterns were generated. ECIZ4 most closely approximat ed a normal distribution, showing less skewness and kurtosis than z, a nd w. The ECIZ4 cutoff values were affected less by test length and th e IRT model than were z, and w. In contrast, the distribution of w was the least stable over replications, and its cutoff values varied grea tly depending on the IRT model and test length.